The work described in this collaborative research proposal has as its primary aim the development of an understanding of the interaction of the important anticancer drug paclitaxel (Taxol) and the promising anticancer drug epothilone B with their common microtubule receptor, and to use the information derived from this study in the design and development of improved analogs of paclitaxel and possible also epothilone B. The interaction of paxlitaxel and tubulin is believed to be crucial to the anticancer activities of the drugs, and thus knowledge of the details of this interaction will facilitate the design and development of improved analogs. The approach adopted will be to prepare paclitaxel and epothilone B analogs of Virginia Polytechnic Institute and State University that are labeled either with a fluorescent probe or with a stable isotope labeled probe. The fluorescent probes will be used at the State University of New York at Binghamton to elucidate the nature of the paclitaxel-tubulin and epothilone-tubulin interaction through fluorescence spectroscopy, leading to information about the dielectric nature of the fluorophore environment, the fluorophore's accessibility to solvent and the size, shape, and associative behavior of the tubulin macromolecule. The relative orientations of the drugs within the receptor site will be deduced from fluorescence resonance energy transfer measurements. The stable isotope labeled analogs will be analyzed in solid state NMR experiments at Washington University using sophisticated phasing experiments such as double-REDOR NMR to obtain information on key internuclear distances of paclitaxel and of epothilone B bound to microtubules. The information derived from both approaches will be combined with molecular modeling information at SUNY Binghamton to develop a model for the conformation of paclitaxel in the binding site of polymerized tubulin. In addition, the best available information on the likely three-dimensional structure of tubulin will be combined with the paclitaxel model into a combined model for the nature of binding site. These models will then be used as the basis for synthetic efforts designed to prepare analogs that are conformationally restricted so as to mimic the active site conformation. Such analogs are expected to show improved bioactivity as compared with paclitaxel itself. In addition, new analogs of paclitaxel will be prepared by combinatorial chemical methods.